Regulators and healthcare providers have stepped up their games to identify drug and device safety risks, capitalizing on growing amounts of electronic patient records and claims data to track adverse events and avert disaster. At the FDA, for instance, the Sentinel System has been taking shape in recent years to collect data on safety snafus related to approved health products and enable regulators to take quick action.
In the past, the FDA relied on doctors and others to report safety problems and the quality of the data was spotty. Plus, who knew whether some of the really bad stuff was even being reported to the agency's Adverse Event Reporting System? The Sentinel program's pilot, dubbed Mini-Sentinel, features a database with aggregated drug safety info on some 126 million people and billions of encounters. And rather than rely on individuals to submit the reports, the pilot system actively monitors info from its data providers' scheduled submissions.
Analytics and modeling offer the chance to prevent deaths from adverse events and other risk factors. For instance, GNS Healthcare revealed last year that it's working with Brigham and Women's Hospital in Boston to develop predictive models of adverse drug reactions and hospital readmissions for patients with congestive heart failure, providing supercomputing-enabled tools to stave off awful outcomes. In nearby Cambridge, MA, incidentally, Novartis ($NVS) scientists have applied computer models built with collaborators at the University of California-San Francisco with troves of side-effect data to predict whether a compound will cause bad reactions before it enters clinical trials.